- Status
I think people need to think more about what nintendo needs instead of what they can get.
Nintendo performance on both wii-u and 3DS where rumored to be far higher then they really where. I believe the wii-u would be 500 gflops gpu performance, while ended up with 176 gflops at the end ( almost a generation below the xbox one, and a generation below the ps4 ).
The 3DS same story, 40-60m poly's ended up being a 12m poly's chip. Both came not even remotely close towards what was rumored.
We know the dev kits has a X1 in it, what clocks it runs at nobody knows. So we can only speculate. And people tend to do exactly what they did with nintendo in the past. speculate on what is possible in relation towards the competition, which frankly is the wrong way to look at it.
What the hardware can do isn't interesting towards nintendo, it's what they need it to do is what matters to them.
With a 500 gflop 25gb/s tablet, multiplatform ports are already a thing of the past. they clearly never ment to get that going with this device. Otherwise they would have delivered a beast of a console that can compete with xbox/ps4 which is easily made so many years after there releases.
It feels exactly the same as people that talked about the wii-u that the chip inside of it probably is overclocked etc etc, so it pushes 700 gflops and competes at 720p vs xbox one at 1080p, while in reality it doesn't even come remotely close towards it ).
So it's safe to assume they are not banking on 3th party developers, but they are banking on there own developers. Now you probably wonder why because they already did this with the Wii-U and 3DS?
Yes they did. But they never combined the two.
That's the issue that sony also had with the psp and vita, they simple can't support both platforms. Otherwise both platforms will be starved for games instead of only just one. This has been proved over the years already. It's probably a struggle that nintendo also had to under go throught with the 3DS and Wii-U, as development of started to become bigger and longer with more graphical games. It's clearly visible tho.
Basically it comes down to, the Switch needs to run wii-u titles ( as it showcased wii-u games ) at it's screen resolution 720p, and at a minimum of 30 fps just like the wii-u. Everything above that is luxery really. This means that nintendo can have a chip in the Switch that doesn't push more then 200 gflops. A heavily underclocked chip to save power with new technology. Which gets boosted to 400 gflops ( 1080p ) for the tv screen. Probably also easy to make for nvidia.
That's what is needed as absolute minimum.
And in my vision that's exactly what is going to happen most likely.
Now ofcourse they can push for more detail in games, higher framerate etc etc. and upclock the gpu to whatever they want it to do. But at the end of the day, they only kill battery time with it, without much result. As they are in compleet control on what the wii-u needs to output.
So instead of looking what is possible from the chip point of view, people need to look at it the way nintendo does.
This is most likely also the reason we will probably never hear about the chip performance from nintendo themselves. And heavily NDA it, because the moment word goes out. It will kill people's hype instantly as the only thing they will look at is the numbers. And without those they can keep on staying in there dream of it being something more then it really is.
If they would push hardware big time forwards, with high clocks and game changing performance, they would scream it off the roof tops like sony did.
Nintendo performance on both wii-u and 3DS where rumored to be far higher then they really where. I believe the wii-u would be 500 gflops gpu performance, while ended up with 176 gflops at the end ( almost a generation below the xbox one, and a generation below the ps4 ).
The 3DS same story, 40-60m poly's ended up being a 12m poly's chip. Both came not even remotely close towards what was rumored.
We know the dev kits has a X1 in it, what clocks it runs at nobody knows. So we can only speculate. And people tend to do exactly what they did with nintendo in the past. speculate on what is possible in relation towards the competition, which frankly is the wrong way to look at it.
What the hardware can do isn't interesting towards nintendo, it's what they need it to do is what matters to them.
With a 500 gflop 25gb/s tablet, multiplatform ports are already a thing of the past. they clearly never ment to get that going with this device. Otherwise they would have delivered a beast of a console that can compete with xbox/ps4 which is easily made so many years after there releases.
It feels exactly the same as people that talked about the wii-u that the chip inside of it probably is overclocked etc etc, so it pushes 700 gflops and competes at 720p vs xbox one at 1080p, while in reality it doesn't even come remotely close towards it ).
So it's safe to assume they are not banking on 3th party developers, but they are banking on there own developers. Now you probably wonder why because they already did this with the Wii-U and 3DS?
Yes they did. But they never combined the two.
That's the issue that sony also had with the psp and vita, they simple can't support both platforms. Otherwise both platforms will be starved for games instead of only just one. This has been proved over the years already. It's probably a struggle that nintendo also had to under go throught with the 3DS and Wii-U, as development of started to become bigger and longer with more graphical games. It's clearly visible tho.
Basically it comes down to, the Switch needs to run wii-u titles ( as it showcased wii-u games ) at it's screen resolution 720p, and at a minimum of 30 fps just like the wii-u. Everything above that is luxery really. This means that nintendo can have a chip in the Switch that doesn't push more then 200 gflops. A heavily underclocked chip to save power with new technology. Which gets boosted to 400 gflops ( 1080p ) for the tv screen. Probably also easy to make for nvidia.
That's what is needed as absolute minimum.
And in my vision that's exactly what is going to happen most likely.
Now ofcourse they can push for more detail in games, higher framerate etc etc. and upclock the gpu to whatever they want it to do. But at the end of the day, they only kill battery time with it, without much result. As they are in compleet control on what the wii-u needs to output.
So instead of looking what is possible from the chip point of view, people need to look at it the way nintendo does.
This is most likely also the reason we will probably never hear about the chip performance from nintendo themselves. And heavily NDA it, because the moment word goes out. It will kill people's hype instantly as the only thing they will look at is the numbers. And without those they can keep on staying in there dream of it being something more then it really is.
If they would push hardware big time forwards, with high clocks and game changing performance, they would scream it off the roof tops like sony did.
Last edited:
Benetanegia
Regular
Wins 3 out of 4. And only loses OpenSSL which afaik, is not really dependant on CPU performance, but rather especialised hardware.
And of course this can only help the A57. I don't think so, it can go both ways, so I don't know why people keep bringing this kind of argument.
The link also shows C-Ray which is a realistic task. FFTE performance is consistent with real world tests when compared to other CPUs as seen here (i.e when compared to Celeron J1900):
http://www.anandtech.com/show/8067/...athlon-53505150-and-sempron-38502650-tested/3
Why would I compare A57 to anything but the CPU on the consoles when the topic is if it could be an alternative to them?
It was already established earlier in the thread that Jaguar was the only low power 64-bit CPU available at the time, seemingly making it the only alternative. That includes A57. It was argued that A57 would not be fast enough anyway, which is what I'm arguing against.
Yes.Measured against the competition in the same price bracket, it's low. Within those limits, devs will make great looking games, no doubt. But with more power they could do more. Taking your Wii U examples, if Wii U had twice the power it could play the same games with far better IQ. Most importantly, it'll be able to play games from other developers that don't limit themselves to simpler art styles, adding considerable value to the platform.
I'm not aware of the competition offering a hybrid console that can potentially play modern games / engines on the go. It's all relative
.Even if the NS GPU is 'only' 500gflops with 'only' 3GB's of RAM for games, people need to realise that this isn't a WiiU situation again for a variety of reasons. The NS GPU supports all modern engines so if a publisher sees an opportunity to make some cash by porting their big name AAA game to it then they will do it even if it means their games are downgraded to equivalent 'low' PC settings and run at 720p instead of 900/1080p on PS4/XB1. If people still don't think the latest AAA games will be possible on NS they need only look at Forza Horizon 2 and Rise of the Tomb Raider on Xbox 360. Both phenomenal 'downports' to a system 5-10x weaker than Xbox One which both hold up incredibly well indeed despite being farmed out to third party development houses.
People comparing NS to PS360 instead of PS4/XB1 are in for a rude awakening when Nintendo show games early next year
.
Last edited:
phoenix_chipset
Regular
It's going to be more powerful than Wii U, which is all I care about. Besides some image quality deficits, Wii U graphics are pretty good. 3x Wii U while docked would be a very good boost. And the cpu will be like a generation and a half ahead.
Although I saw some aliasing in the new Mario footage, kinda looks like 720p which has me a bit worried.
Although I saw some aliasing in the new Mario footage, kinda looks like 720p which has me a bit worried.
The safest bet for my own personal sanity is to expect performance exactly in line with what Eurogamer leaked, the Tegra X1. Seeing as how they were dead on with their info, I see no reason to doubt the X1 is very much inline with the Tegra chip powering Switch. We know its custom, but that could be modifications designed around power efficiency and perhaps removing parts such as the A53 cores. Nvidia in their blog talked about all the man hours that went into this, but they talk more about the API and tools than hardware, which makes me believe that more time went into creating tools and a very efficient lightweight API to get the most out of the Tegra processor.
What you write sounds logical, but after the Wii and before Wii U Nintendo was in a unique position. People bought the Wii despite it outdated specs because it had a USP: Motion control. With the Wii U they thought they were in the same position because of the Wii U GamePad. But people did not bite this time and with the novelty of the motion control feature waning the Wii U looked pretty bad compared to XB1 and PS4. And it was a bitch to develop for because of the under powered CPU. I remember how a lot of developers and publishers jumped ship very early, because of the additional effort required for ports. In the end Nintendo sold a lot less consoles than they had anticipated. The shareholders were not happy.
If history repeats itself with the Switch and it ends up a "least effort" console who's gonna buy it? Especially with PS4 Pro out now and XB1 Scorpio available next year. It don't think the mobility feature is enough of a USP to make people forget about hopelessly dated graphics. Plus mobile phones get better every year as well.
If Switch is not way more successful than the Wii U, it will be the last (stationary or semi-stationary) console by Nintendo. The shareholders will be even less happy and that means people in high places losing jobs.
So if Nintendo really goes the "least effort" route like you predicted they need to be damn sure they can make the Switch a successful never less. Maybe there are other killer features we have not seen yet, but I doubt it.
Another variable in the equation is nVidia's pride. I would not underestimate that.
BTW: The game footage in the trailer was probably inserted in post-production, so we cannot really be sure what to expect by watching the trailer: http://www.eurogamer.net/articles/2016-10-24-behind-the-scenes-of-nintendos-switch-reveal-video. It is possible that there is no final hardware yet.
I think a safe bet is to assume Shield performance.
Reading reviews of the Shield, there's mention of the fan being very quiet and inaudible 15cm away, and not running unless under load, as expected.
Process change or hardware update (Pascal ?), could improve performance and/or autonomy ^^ (I'd bet on the second for battery life and first once docked for 1080)
Reading reviews of the Shield, there's mention of the fan being very quiet and inaudible 15cm away, and not running unless under load, as expected.
Process change or hardware update (Pascal ?), could improve performance and/or autonomy ^^ (I'd bet on the second for battery life and first once docked for 1080)
Course it is. "Low" is a relative comparison where the benchmark needs to be specified. NS is good enough for a handheld, no doubt. When it's docked, it's low spec for high-fidelity TV console. Obviously for those who want console games on the go with TV output as a secondary function, NS is probably good enough.I'm not aware of the competition offering a hybrid console that can potentially play modern games / engines on the go. It's all relative
Which people are saying, thinking, or doing any of that?...people need to realise that this isn't a WiiU situation again...If people still don't think the latest AAA games will be possible on NS...People comparing NS to PS360 instead of PS4/XB1 are in for a rude awakening when Nintendo show games early next year
Last edited:
In a handheld, maybe. But seriously, where everyone else manages a console generational advance of ~8x, creating a vastly and notably superior experience, why should 3x after 5 years from a fairly underpowered machine be acceptable for a TV machine?
When I write "final hardware" I mean the actual non-prototype, non-devkit hardware. Look at these leaked kaby lake roadmaps: https://benchlife.info/intel-will-launch-18w-tdp-kaby-lake-h-in-2017-q2-1022016/
The mobile kaby lake H CPUs apparently go into production week 43 2016 and are ready to ship around week 51 or 52. That's just 8 or 9 weeks. If the switch launches in japan in march 2017, the actual hardware may have gone into production after the trailer was filmed and is not ready yet.
Last edited:
But then it needs to be soldered, Switch need to be produced, packed, shipped...
Maybe Christmas will be enough for all that, as I said I'm no specialist, it just sounds risky.
Compared with the 4.5W TDP A10 Micro 6700T for fanless tablets (isn't that the slightly better comparison here?) it loses 2 out of 4. Calculating the geometric mean of the four results has the Mullins chip 48% in front of the Shield. Mullins is almost a year older, has only 10.6GB/s memory bandwidth, is produced on 28nm and features a 4.5W TDP while the Tegra X1 of the Shield is produced in 20nm, has 25.6GB/s at its disposal and consumes "less than 10W". Considering that, they are probably roughly in the same league with too few benchmarks around to decide for sure.
TLDR: I wouldn't blindly assume a quad A57 is in front of a quad Mullins all other things being equal.
Last edited:
phoenix_chipset
Regular
Very good in the context of how slowly current tech is evolving, and that Nintendo can do a lot with a little.
Also Xbox one certainly wasn't 8x 360, and both the ps4 and xb1 saw like a 3x cpu boost for games over the previous generation.
I'm no expert either but don't phone manufactures do that all the time? According to reports mass production of iPhone 7 started in June 2016 and it hit streets in September. That's just 3 months. If Switch will first be released in Japan at least shipping should not be the issue.
Technology is still progressing at a fair rate, especially from Nintendo's console design. That is a 4 TF console at a mainstream price is perhaps what would be expected of a new machine and a generational advance on Wii U. Obviously NS is making concessions to hit a portable form factor, but that does mean as a TV console it's very underpowered.
Meaning? They aren't any better at extracting performance than anyone else. They just focus on different targets, and they're not afraid to short-change their consumers. By that I mean the likes of Wii which was 'good enough' to realise the game experiences, but actually pretty crap graphics with really low IQ when far better was possible. I guess if Nintendo fans are happy to accept shimmers and jaggies and blurry IQ on upscaled games, Nintendo will feel no pressure to provide more capable hardware...
'8x' isn't meant as a sane metric. How does one measure the power advance? Increase in flops alone? Increase in BW? Lowest increase of any part of the system (weakest link) and greatest individual component increase? As a ball park, '8x' is a general summary, illustrative figure, covering the Moore's Law expectation over 6 years. Regardless of what measurement we use, on screen results are a visible generational advance from one console to the next for every console manufacturer except Nintendo. GC > Wii looked the same, and Wii U > NS is likely going to be pretty similar (again, because it's a handheld and not a TV console! So understandable, but still true when comparing TV consoles).
I can't really think of any GC game that looks as good as Mario Galaxy 1+2, or Skyward Sword... I say this having played a shit ton more GC games than Wii(barely used the damned thing. Would have used it a lot more if more games supported controllers that actual crippled people can use...)
Wii isn't a generational advance because it's the same hardware! Nintendo just learnt some more tricks with it along with more headroom from the higher clocks. Go look up what the demoscene is able to achieve on the ZX Spectrum (Timex 2000) 30 years after release to see how hardware can be pushed!
- Status
Similar threads
